1. Field of the Invention
This invention relates to agents and conjugates used in the detection and isolation of targets from heterologous mixtures. Agents comprise a detectable moiety bound to a photoreactive moiety. Conjugates comprise agents which are coupled to substrates by one or more covalent bonds. These bonds can be easily and selectively cleaved or photocleaved with the application of electromagnetic radiation. Substrates which may be coupled to agents include amino acids, peptides, proteins, nucleotides, nucleic acid primers for PCR reactions and lipids. The invention also relates to rapid and efficient methods for the detection and isolation of targets, such as cells, nucleic acids and proteins, and to kits which contain these components.
2. Description of the Background
Basic scientific techniques including some of the major breakthroughs in molecular biology, chemistry and medicine have certain features in common. Two of these features are the specific detection and isolation of individual components from complex mixtures. For example, electrophoresis and chromatography are each widely utilized procedures to detect or isolate macromolecules from biological samples. These procedures take advantage: of unique or identifiable molecular properties of the components to be isolated such as charge, hydrophobicity and molecular weight, to characterize and identity macromolecules. Depending on their method of isolation, macromolecules isolated can often be utilized as products in downstream processes.
Some of the more useful detection and isolation procedures take advantage of physical properties of the element of interest, the substrate, or of molecules which can be easily attached to substrates. One of the most widely used of these properties is radioactivity and radioactive labeling with radionuclides. For the most part, substances are not naturally radioactive and can be labeled with radioactive atoms, referred to as radionuclides, and detected using standard and well-known radiographic procedures. Radioactive elements are detectable because they emit large amounts of energy in the form of alpha, beta or gamma rays as they decay. Radioactivity is generally useful for labeling because the label is not affected by the physical state or chemical combination of the substance to be labeled. In addition, the specific radiation emitted can be identified by the nature of the radiation (e.g. α, β or γ), its energy and the half-life of the process. Targets can be identified in complex mixtures from the radiation profile emitted. Further, radioactively labeled substances can be followed radiographically in chemical pathways and biological systems.
Unfortunately, radioactivity is a hazard to both human health and the environment. The protection which must be afforded each worker is substantial. Special laboratory procedures, dedicated facilities and equipment, detailed record keeping and special training of laboratory personnel are all required for the safe use of radionuclides. Production of radioactive reagents is also very expensive as is safe disposal which drives up the cost of all experiments involving radioactive agents. Further, under present guidelines, all users of radioactivity require specialized supervision and federal regulations must be strictly and carefully adhered to requiring an enormous amount of record keeping.
Radioactive labeling methods also do not always provide a means of isolating products in a form which can be further utilized. The presence of radioactivity compromises utility for further biochemical or biophysical procedures in the laboratory and in animals. This is clear in the case of in vitro or in vivo expression of proteins biosynthetically labeled with radioactive amino acids or tagged with other radioactive markers. The harm or at least potential harm of the radioactivity outweighs the benefits which might be produced by the protein composition.
Disposal of radioactive waste is also of increasing concern both because of the potential risk to the public and the lack of radioactive waste disposal sites. In addition, the use of radioactive labeling is time consuming, in some cases requiring as much as several days for detection of the radioactive label. The long time needed for such experiments is a key consideration and can seriously impede research productivity. While faster methods of radioactive detection are available, they are expensive and often require complex image enhancement devices.
There are many other detectable physical properties which can exist in chemicals and chemical moieties that can be used to detect and isolate substances. One of these physical properties is the property of luminescence which includes the phenomena of fluorescence and phosphorescence. Fluorescent chemicals emit radiation due to the decay of the molecule which has been excited to a higher electronic state due to the absorption of radiation. Phosphorescent molecules can emit radiation for a much longer time intervals. Detection of the specific wavelength of radiant energy emitted allows for the detection of targets which may be associated with the luminescent chemicals.
Bioluminescence is rapidly becoming a widely used method for labeling many different types of compounds. Basically, a reduced substrate is reacted with oxygen and converted into an oxidized product with an elevated or excited electronic state. The excited molecule decays to the ground state and in the process, emits photons of light. This process has been found to occur in several strains of bacteria and fungi in marine invertebrates such as sponges, and in shrimp and jellyfish. Bacteria which emit light are often found living symbiotically with fish in special luminescent organs. A wide variety of terrestrial organisms such as earthworms, centipedes and insects also possess bioluminescent properties.
One group of compounds which undergo oxidation with the emission of light are referred to as luciferins although their individual structure may vary. The oxidized products are termed oxy-luciferins and the enzymes which catalyze the process luciferases. The overall process is endothermic requiring chemical energy stored in one to two molecules of adenosine triphosphate (ATP) per photon of light produced. Two types of luciferase systems that have been widely used in molecular biology are the bacterial system (Vibrio haveyi or V. fischen) and the firefly system (Photinus pyralis).
Other labels which impart detectable properties to a substrate include chemicals with a unique absorption spectrum, electron spin resonance spectrum, optical activity, Raman spectrum or resonance Raman spectrum. Such labels are widely used in many fields including medicine, molecular biology and chemistry. For example, the visible or infrared absorption spectrum of a molecule often constitutes a unique fingerprint which allows the molecule to be identified even in the presence of a complex mixture. In the case of visible absorption, the molecule absorbs radiant energy over a specific wavelength range because of the presence of an excited electronic state of the molecule whose energy of transition from the ground state falls in the range 1.5-3 eV. In the case of infrared absorption, bands are detected due to the excitation of vibrational modes of the molecule. The frequency of these bands provides information about the presence or absence of characteristic molecular groups such as disulfides, carbonyls and aromatic groups.
In another application of the spectroscopic properties of molecules, nuclear magnetic resonance (NMR) spectroscopy has been extensively used to identify specific molecules in a mixture. The nuclei of atoms, such as protons in hydrogen atoms, which possess a net magnetic moment, will align when placed into a magnetic field with that field and will precess about that field with a frequency (the Lamar frequency) dependant on the individual properties of the particle. To determine the NMR spectrum, a sample of protons is placed within a strong magnetic field and irradiated with a range of radio frequencies at a 90° angle with respect to the main field. This treatment causes all the protons in the sample to absorb energy at their characteristic frequency, flipping their magnetic orientations 90° with respect to their original state. After the applied field is switched off, the molecules gradually relax to precess about the main field. Receiver coils which surround the sample detect the frequencies of precessing spins as a set of oscillating electric currents which constitute the NMR signal.
All of these methods suffer from similar disadvantages. For the most part, targets do not have unique detectable properties such as inherent radioactivity or fluorescence. Labels must be attached which are themselves detectable and therefore make the target detectable. However, the labeling process can result in a labeled product that is in some way permanently damaged. For example, fluorescent chemicals can be extremely toxic to cells. Long term exposure can result in a high degree of cell death. Often, the labeling compound may have detrimental effects on a target's structure or activity. Protein structure is often adversely affected by the attachment of a detectable chemical moiety. Labeling of nucleic acids can interfere with their ability to be translated, transcribed by polymerases or interact with DNA binding proteins. In most cases, the chemical moiety must be removed. Further, the methods for removal of these chemical moieties which have detectable physical properties often result directly in alteration of the molecule or cell death.
There are a number of procedures, both complex and simple, which have been used to selectively detect and isolate target substrates. One procedure which has revolutionized and greatly accelerated the detection and identification of nucleic acids is polymerase chain reaction (PCR) technology. The principle concept of PCR is the rapid, large-scale amplification of unique or even non-unique nucleic acid sequences in biological samples. Using labeled primers with specific or random sequences, the genetic code of very small quantities of nucleic acids can be detected, amplified in number and subsequently characterized through repetitive polymerization events. Although the nucleic acids formed are new, the sequence of the original sample is maintained and can be easily determined. As a nucleic acid sequence is the biological code for the construction of virtually all proteins, the origin, evolutionary age, structure and composition of nearly any biological organism or sample can be determined from knowledge of the sequence. The procedure has been proved useful in molecular and evolutionary biology, and has demonstrated applications in the detection, treatment and prevention of diseases and disorders in humans.
PCR technology, although revolutionary, carries with it the same limitations as many conventional detection and isolation procedures. Label which has been incorporated into primers and ultimately newly formed nucleic acids must be removed. This process, when possible, is fairly time consuming and often results is modification or destruction of the nucleic acid.
Another example of a process to render a substance specifically detectable is to use binding molecules which have a particular affinity for selected other molecules as occurs between binding of an antigen to an antigen-specific antibody. These chemical pairs, sometimes referred to as coupling agents, have been used extensively in detection and isolation procedures. Normally one of the molecules in this pair is immobilized on an affinity medium such as used in chromatographic packing material or a magnetic bead and used in the isolation of the target molecule. Some of the more useful coupling agents are biotin and avidin or the related protein, streptavidin. These agents have been used in many separation techniques to facilitate isolation of one component or another from complex mixtures
Biotin, a water-soluble vitamin, is used extensively in biochemistry and molecular biology for a variety of purposes including macromolecular detection, purification and isolation, and in cytochemical staining. Biotin also has important applications in medicine in the areas of clinical diagnostic assays, tumor imaging and drug delivery, and is used extensively in the field of affinity cytochemistry for the selective labeling of cells, subcellular structures and proteins.
Biotin's utility stems from its ability to bind strongly to the tetrameric protein avidin, found in egg white and the tissues of birds, reptiles and amphibians, or to its chemical cousin, streptavidin, isolated from the bacterium Streptomyces. Typically, biotin or a derivative of biotin is first bound directly to a target molecule, such as a protein or oligonucleotide, or to a probe using specific chemical linkage. The interaction of the linked biotin with either streptavidin or avidin conjugated to an affinity medium such as magnetic or sepharose beads is then used in the isolation of the target molecule. Alternatively, the interaction of the covalently linked biotin with avidin or streptavidin conjugated to an enzyme such as horseradish peroxidase (HRP) which catalyzes a chromogenic reaction is used for detection of the target molecule. Macromolecules that have been isolated using biotin-avidin technology are shown in Tables 1 and 2.
TABLE 1Macromolecules Isolated by Direct BiotinylationBiotinylatedTargetsElution ConditionsReferencesMembrane proteinsacetate, pH 41, 2and glycoproteinsAntibodieslow pH3Enzymesnon-physiological4t-RNA6M guanidine-HCl,5pH 2.5rRNA70% formic acid6nucleosomesSS-reduction of7, 8cleavable biotinDNAnon-physiological91. G. A. Orr, J. Biol. Chem. 256:761, 1981.2. C. A. Beard et al., Mol. Biochem. Parasitol. 16:199, 1985.3. E. A. Bayer et al., FEBS Lett. 68:240, 1976.4. C. S. Chandler et al., Biochem. J. 237:123, 1986.5. T. R. Broker et al., Nucl. Acids Res. 5:363, 1978.6. D. J. Eckermann et al., Eur. J. Biochem. 82:225, 1978.7. M. L. Shimkus et al., Proc. Natl. Acad. Sci. USA 82:2593, 1986.8. M. L. Shimkus et al., DNA 5:247, 1986.9. P. L. Langer et al., Proc. Natl. Acad. Sci. USA 78:6633, 1981.
TABLE 2Biological Materials Isolated Using Biotinylated Binding MoleculesBindingTarget MoleculesMoleculesElution ConditionsRefs.Glycoproteinsconconavalin A2% SDS10Membrane AntigensantibodySDS (boiling)11Estrogen Receptorestradiacetate,estradiol12Insulin Receptorinsulinacetate, pH 5.013, 14biotinOpoid Receptorenkephalinenkephalin15Human Bantigenselection by FACS16lymphocytesLymphocytemonoclonalMechanical17, 18, 19subpopulationsantibodyagitation,erythrocyte lysisPlasmid DNADNA0.1 M NaOH20SpliceosomesRNA90° C. in SDS21RecombinantDNACleavable biotin22PlasmidsHeat, low ionicstrength and phenol10. J. W. Buckie et al., Anal. Biochem. 156:463, 1986.11. T. V. Updyke et al., J. Immunol. Methods 73:83, 1984.12. G. Redeulih et al., J. Biol. Chem. 260:3996, 1985.13. F. M. Finn et al., Proc. Natl. Acad. Sci. USA 81:7328, 1984.14. R. A. Kohanski et al., J. Biol. Chem. 260:5014, 1985.15. H. Nakayama et al., FEBS Lett. 208:278, 1986.16. P. Casali et al., Sci. 234:476, 1986.17. J. Wormmeester et al., J. Immunol. Methods 67:389, 1984.18. P. J. Lucas et al., J. Immunol. Methods 99:123, 1987.19. R. J. Berenson et al., J. Immunol. Methods 91:11, 1986.20. H. Delius et al., Nucl. Acids Res. 13:5457, 1985.21. P. J. Grabowski et al., Sci. 233:1294, 1986.22. B. Riggs et al., Proc. Natl. Acad. Sci. USA 83:9591, 1986.
While the utility of biotin continues to grow, there still exists major drawbacks in the use of biotin-streptavidin technology for many applications. This problem stems from the high affinity between biotin and streptavidin, precisely the molecular characteristic which makes it most useful. Once a target molecule or cell is isolated through the streptavidin-biotin interaction, release of the target molecule requires disruption of this interaction. Dissociation of biotin from streptavidin requires very harsh conditions such as 6-8 molar (M) guanidinium-HCl, pH 1.5. Such conditions also denature, and thereby inactivate, most proteins and destroy most cells.
For example, a biotin derivative containing a N-hydroxysuccinimide ester group is commonly used to link biotin through an amide bond to proteins and nucleic acids. Selective cleavage of this linkage disrupts similar native chemical bonds in associated molecules. Biotin is also often used in the isolation of specific cells from a heterogeneous mixture of cells by binding a biotinylated antibody directed against a characteristic cell surface antigen. The interaction of the biotinylated antibody with streptavidin-coated magnetic beads or sepharose particles can then be used effectively to isolate target cells. Disruption of the antibody-antigen interaction normally requires exposure of cells to conditions such as low pH or mechanical agitation which are adverse to the cell's survival. In general, recovery of the target in a completely unmodified form is not possible.
Once the biotinylated DNA is bound to streptavidin it can only be released with extreme difficulty. Many diverse methods to remove the streptavidin molecule have been suggested including digestion by proteinase K (M. Wilchek and E. A. Bayer, Anal. Biochem. 171:1, 1988). Proteinase K also digests nearby proteins and does a fairly poor job of completely digesting the streptavidin. Significant amounts of the streptavidin molecules remain attached, and further, removal of streptavidin does not release the biotin. Further, biotinylated DNA interferes with subsequent use in a variety of methods including transformation of cells and hybridization based assays used for detection of genetic diseases.
The essentially irreversible binding of biotin and streptavidin is also a serious limitation for the performance of multiple or sequential assays to detect a specific type of biomolecule, macromolecular complex, virus or cell present in a single sample. Normally, only a single assay can be performed because the enzyme detection system is streptavidin-based and streptavidin remains firmly bound to the biotinylated target or target probe. While different chromogenic systems for detection are available, they are only of limited applicability in situations where large numbers of probes are needed.
An additional problem in the use of biotin-avidin technology is the presence of endogenous biotin, either free or complexed to other molecules, inside the sample to be purified or assayed. In this case, the endogenous biotin can result in the isolation or detection of non-target molecules. This can be a particularly severe problem in cases where a high signal-to-noise ratio is needed for accurate and sensitive detection.
To remove biotin from an attached molecule, several chemically cleavable biotin derivatives have been produced. Immunopure NHS-SS-biotin (Pierce Chemical; Rockford, Ill.) consists of a biotin molecule linked through a disulfide bond and an N-hydroxysuccinimide ester group that reacts selectively with primary amines. Using this group, NHS-SS-biotin can be linked to a protein and then the biotin portion removed by cleaving the disulfide bond with thiols. This approach is of limited use since thiols normally disrupt native disulfide bonds in proteins. Furthermore, the cleavage still leaves the target cell or molecule modified since the spacer arm portion of the complex is not removed and the cleaving buffer must be eliminated from the sample.
One method for removal of biotin is the use of disulfide-based cleavable biotins. However, the cleaved molecules possess a reactive sulfhydryl group which has a strong tendency to form disulfide bonds with other components of the mixture. Functional activity of these substances containing sulfhydryl groups is severely compromised. Typically, activity of such protein is decreased or eliminated and such nucleic acids will no longer hybridize rendering them useless for cloning. This method is also slow and requires the preparation of complex solutions.
An additional limitation of biotin-avidin technology is the difficulty of developing automated systems for the isolation and/or detection of targets due to the problems of releasing the target from the biotin-avidin binding complex. This requires addition of specific chemical reagents and careful monitoring of the reactions.
Biotin-avidin technology has been combined with PCR techniques for the detection and isolation of nucleic acids and specific sequences. However, there still remains a fundamental problem which relates to the difficulty of removing the incorporated biotin. This is normally not possible using conventional biotins without irreversibly altering the structure of the DNA. As discussed, biotinylation can interfere with subsequent application of biotinylated probes as well as alter the properties of the PCR product.
PCR products that contain biotinylated nucleotides or primers which are required for isolation cannot be used in conjunction with biotinylated hybridization probes. The presence of biotin on the PCR product cause false signals from the avidin based enzyme-linked detection system. Biotin incorporation into DNA interferes with strand hybridization possibly due to the spacer arms linking the nucleotides to the biotin molecules. Further, PCR products that are biotinylated are not suitable material for cloning. PCR products which contain biotinylated nucleotides are difficult to analyze. Incorporation of biotinylated nucleotides into DNA causes a retardation of mobility during gel electrophoresis in agarose. This mobility shift renders characterization of PCR products difficult. As proper DNA-DNA hybridization is the basis for sensitive and accurate characterization and sensitive assays, biotin-avidin binding systems are seriously disadvantaged.
Other coupling partners which can be used to detect and isolate target substances are cell adhesion molecules (CAMs). One of the well characterized types is the endothelial cell adhesion molecule, LEC-CAM (leukocyte endothelial cell-cell adhesion molecule), now called selectin. This molecule selectively binds to leukocytes. Its natural function is to facilitate the transport of leukocytes through an endothelial layer of cells such as postcapillary venules to sites of inflammation or tissue damage. There are many of these adhesion molecules which have been identified in humans and other mammals that range in binding specificity from the very general to the highly specific. These include the endothelial cell adhesion ligands ICAM-1, VCAM-1 and ELAM-1, the β-integrins which consists of a family of three proteins LFA-1, Mac-1, VLA-4, MO-1 and p150/95, carbohydrate binding CAMs that appear on endothelial cells, platelets, and leukocytes, and the cadherins, calcium dependent CAMs present on most cells. Attachment of these molecules or the creation of fused proteins containing adhesion domains can be used to facilitate isolation and detection of binding partners. However, once binding has occurred, complex, expensive and time consuming biochemical manipulations and sometimes fairly harsh chemical treatments are necessary to dissociate the molecules. Further, application of these molecules for general use is limited as binding partners must be located for each target of interest.
Other coupling partners include nucleic acids and nucleic acid binding proteins, lipids and lipid binding proteins, and proteins or specific domains which have a particular affinity for each other. These coupling partners suffer from similar drawbacks as the biotin-avidin system and the adhesion molecules.
Another fairly ubiquitous method of detection and isolation is gel electrophoresis. In this process, a uniform matrix or gel is formed of for example, polyacrylamide, to which is applied an electric field. Mixtures applied to one end of the gel will migrate through the gel according to their size and interaction with the electric field. Mobility is dependent upon the unique characteristics of the substance such as conformation, size and charge. Mobilities can be influenced by altering pore sizes of the gel, such as by formation of a concentration or pH gradient, or by altering the composition of the buffer (pH, SDS, DOC, glycine, salt). One- and two-dimensional gel electrophoresis are fairly routine procedures in most research laboratories. Target substances can be purified by passage through and/or physical extraction from the gel.
Methods for the detection and isolation of targets substances also include centrifugation techniques such as equilibrium-density-gradient centrifugation. This process is based on the principal that under high centrifugal forces, stable gradients will be established in salt solutions. Mixtures subjected to high speed centrifugation will segregate individual components according to their specific densities. Although useful, all of these procedures are more quantitative than qualitative.
A major advance in detection and isolation methodology was the advent of liquid chromatography. Chromatography, and in particular column chromatography, comprises some of the most effective and flexible purification methods available. Common to most procedures is the use of open cylinders containing a hydrated matrix material. Some of the typical matrix materials which are presently used, for example, in gel filtration, affinity chromatography and ion exchange chromatography, include sepharose (bead formed gel prepared from agarose: Pharmacia Biotech; Piscataway, N.J.), sephadex (a bead-formed gel prepared by cross-linking dextran with epichlorohydrin: Pharmacia Biotech; Piscataway, N.J.) and sephacryl (covalently cross-linked alkyl dextran with N,N′-methylene bisacrylamide: Pharmacia Biotech; Piscataway, N.J.). Basically, a heterogenous sample or mixture is applied to the top of the column followed with a suitable buffer. Substances within the mixture display differential migration through the column in relation to other materials within the sample and is collected in fractions at the other end of the column. Fractions are individually analyzed for the presence of target and positive fractions pooled.
Alternatively, target in the sample may selectively bind to the column material in the presence of buffer a process known as affinity chromatography. After binding, unbound material is removed by continuously washing the column with buffer. Target molecules are subsequently released from the column by application of an elution buffer which causes dissociation. Fractions are collected as they elute off of the column and collected. In gel-exclusion chromatography, a cross-linked dextran is utilized as column matrix material. Cross-linking can be varied to alter the effective pore size of the column material, and the dextran can be coupled to a wide variety of chemical moieties to selectively capture target. Ion-exchange chromatography takes advantage of the fact that targets, for example proteins, can differ enormously in their affinity for positive or negative charges on column materials. The affinity of a material for a target is proportional to the salt concentration of the buffer. By raising or lowering the salt concentration, it is possible to change affinity of target to column material.
Affinity column chromatography makes use of chemical groups that have special attraction to the targets of interest. For example, enzymes preferentially bind to certain naturally associated cofactors. Column materials with attached cofactors will selectively bind to such target enzymes. Enzyme purification becomes a relatively simple and straightforward matter. In a similar fashion, enzyme-specific antibodies can be coupled either covalently or non-covalently to a column matrix. The unique affinity of an antibody for its target antigen allows for the selective removal of target from a heterologous mixture of substances. Detection and isolation is again a fairly simple matter.
Two relatively well-established procedures, high-performance liquid chromatography (HPLC) including reverse-phase HPLC and size-exclusion HPLC, and the more recent technique fast-performance liquid chromatography (FPLC) which can handle larger sample volumes than HPLC, is based on standard chromatographic techniques, but using extremely high pressures (5,000 to 10,000 psi and more). Due to the higher pressures, finer column materials can be utilized and separations can be performed faster and with better resolution.
Although chromatography is an invaluable tool it too has its limits. Materials to be separated must be solubilized into a suitable buffer which will not adversely affect the column. Further, substrate mixtures and targets must be capable of passing through a column matrix in a reasonable period of time. Although HPLC can sometimes shorten this time period, only small quantities can be detected and the high pressures can damage isolated column material.